 Okay, hi, can everybody hear me? I don't have any feedback on audio and I have no chat I guess you can hear me. So I will answer to do this easiest question That disappeared. Oh, there is 5.10 is the next LTS kernel That's easy. I picked the last one of the year and this one will be in December 20th, I think so that's the last one 5.9 was big and 5.10 is also big. So yes, that's one of the year. That was easy How do I keep up with the ever-growing list for patches for reviews We have scripts that if once a patch gets committed to Linus's tree they pick up the patch and they Let me know that it's there and then because it's already been tagged for stable. I Review it manually Spit it off to the build systems that I have and get supplied and goes from there So it's just we have a lot of backing scripts Sasha has a whole bunch of scripts that does auto Selection of patches and that he digs through all the stuff that lands on Linus's tree and goes from there as well So it is a lot We're averaging about 30 to 35 patches a day. They get merged overall every single day Do you aggregate them all so it's a lot, but we can do it pretty fast I wrote a blog post about how I do a lot of this stuff and how it's automated through mutt and our scripts and all our scripts are Public and mutt stuff is public. So it's pretty easy Other ones this is going upside down Have I tried switching to the Eric email client yet? No, I haven't I haven't looked at it But works good. I was waiting for Drew to say take the time and see if it works for me or not As there are any tips arguments for to convince of the level management to contribute It's very simple It will save you time and money if you contribute if you want to waste time and you want to waste money you do not contribute you everybody has a fork of the tree because they have things they needed to do and You get involved by pushing that stuff upstream that means you don't have to maintain it anymore You get more intimately evolved. It actually has been documented IBM and Intel have publicly written up Articles about this. I think you can Carver business review is written article saying it will save you time and money to work upstream It's that simple. So if your management has lots of money to burn and time to burn Wonderful do it does take a first hit of the first time you get involved and get your patches upstream But over time it is proof that it's faster and cheaper These are upside down Well, I think of the discussion that email patch mythology is holding back contribution That is not what's holding back contributions We have over 200 new developers show up every single release. So every three months, we have 200 new developers We do not have a problem of new developers right now Yes, it is hard to get your email client to work and whatnot, but we have a documented really well And it's proof 200 brand new people show up every month whether that could be 400 brand new people or not Maybe we do need to do that. But and we are working on things like that. We have tutorials. We have posts on how to do this We've been working on Lord a chrome or to make things easier that way But our main bottleneck is maintainers is reviewing. I average I just looked over 700 patches a week that I have to review And that is our bottleneck right now. So Help review Patches help other people if you want to submit a patch, there's no reason why you shouldn't also be reviewing other people's patches It's just like with music. You don't start off writing music only you start off reading music and criticizing music Same thing with programming you should be reading and and Reviewing out of people's code as well some subsystems I know SCSI require this that you have to review other people's code and I think that's good idea So help out the maintainers. That's a really good idea How does the kernel community interact with OSHW movement, I have no idea what OSHW movement is sorry somebody wants to Answer that I will or types that out. I will Does the kernel memory have memory manager for huge pages. Yes, it does we have huge page support So we have a file system for it. We have huge base support. It's been there for 15 20 years. So yes, we have huge base support I do not an MM person. So I'm not the person to ask about that I'm Consideration to be taken when implementing runtime PM callbacks versus system-wide system resume callbacks There's a whole document about all of this in the documentation tree on runtime power management You should do runtime power management. You should not mess with system-wide suspend resume callbacks Because you want to you want to knock off the hardware that is not being used at that moment of time It saves you overall power use that that's what everybody's switching over to there's a document on how to do it It's in a documentation directory. It's much more finer grained. It works much better. I'm pleased to do that Somebody just told me off ban that the OSHW is the open source hardware association How do we okay? So I don't know how the kernel community interacts with the kernel community is 4,000 people in 500 different companies I do not know all of them I'm sure there's some people making sure that the next works on that hardware at Facebook and other companies that work with that So What else does that do oh, thank you drew everybody's just missing that Google is pushing towards a GKI kernel image Yes, do I think it'll encourage mobile vendors to contribute upstream it already has we have proof of that I have seen patches from Samsung and MediaTek and Qualcomm Pushing their core changes to the kernel and drivers upstream. You will see it in a lot of driver subsystems I've already got broken up on IRQ none RQ There's been a number of subsystems that have broken up into being modules now instead of loadable They can be loadable modules. We have patches from MediaTek grabbing from a Qualcomm tree Sending them through an Intel maintainer right now So we have lots of active development happening right now by all these SOC vendors and chip set vendors and phone vendors I work with Google a little bit on this stuff just making sure that the LTS kernels get merged together So yes, we've already seen it contributing upstream very well The problem will SOC mobile vendors put everything in their modules and not care about upstream Everybody does that for all districts. So that's nothing new I'm sure it will continue to happen as well, but there's only so much you can do a module By virtue of what it is and again, it saves companies time and money when they merge stuff upstream So if they merge stuff upstream, they save time of money and they get involved and that's the main goal there It's working pretty well so far Let's see GKI is gonna come out sometime next year. So let's see how it works But then it'll be a while before phones are out on it, but we already have proof that it is helping Was my vice for a little Linux kernel mentees. Oh, hi, this is one of my mentees from the summer How can they be useful after mentorship? They can talk to me, but they can keep doing the same things that they've been doing and Actually, so there's been a number of them I've been trying to work with to figure out what subsystem they want to work on or what they want to specialize in and go What for there? We have one mentor one mentee this summer who actually was offered to be a maintainer of the subsystem It was a little bit of tag here. It you touched it last And they were very surprised, but they are taking up some that tasks right now and doing quite well at it So just get involved and find it's something that you're interested in if you're not interested You still want to generic stuff again, please do the S. Y. Z. What stuff other stuff? We have lots and lots of tasks to do so keep on going that What is a roadmap regarding Linux as a RTOS it's been that way for 10 15 years already the RT patch We're real-time deterministic that operating system It preempt RT is or the real-time stuff is almost all merged. It's not quite there yet I'm talking to share. There's a talk about real-time at this conference. So go ask those developers It is getting whittled down into smaller and smaller bits But it's almost all there and there's a talk at plumbers about what to do when it is all merged in and how do we do testing and integration and stuff Go back and look at the Linux plumbers conference Talks about that for details if you're curious But yeah, it's there you can download it and run. I'm we saw the next running on laser welding robots Decade ago if not older than that They do John Deere the drive tractors. They do lots of other stuff. Linux has been a real-time operating system for a very long time And I ran out of questions What's my take on it? There we go. That's better rust in the kernel. I don't think people have noticed so the rest developers talk to me and Alina's Well, no a year and a half ago and almost two years ago and we said sure wonderful Let's see how it works and they've been actively working to try and get something that will work since then There's some proof of concepts, which is cool if you look at the plumbers conference the rest developers on the kernel developers we talked and Yeah, it looks nice if they can come up with something that will work Well, that's great. Right now you start to run like the bleeding edge rust compiler of every day nightly builds There's some interesting Interactions that's going to happen when you get with object life spans in the sea objects We have versus the rest objects. It'll be interesting to see how they handle that but they're working on it So, yeah, it'll be interesting to see what happens. It's just another language. That's fine So great But if you're more interested and go look up the plumbers conference and there's a whole talk and there's a summary of it on The next week the news as well. So really good How much time do I devote the kernel newbies the IRC channel or the mailing list the IRC channel? I'm on and if there's questions I can see the answer I'll answer it the mailing list if there's questions that come by and I know what to answer I'll answer it. So it's just like one of my normal emails that I get in my inbox I get about a thousand emails. I go do something with every day. That's just one What am I working on now projects wise I'm cleaning up the debug FS interface getting its Smaller and more robust if you look over time I've slowly evolved the API to make it harder to use wrong and then there's a bunch of Make things easier to use with the usb messages Functions as that well as that I worked with some interns on that Getting that the usb api some of them are very easy to use incorrectly As the bad api you want an api to be very hard to use wrong So we changed them and tweaked them a little bit and we'll evolve and move the whole kernel over to the new apis over time So lots of little tiny evolutionary stuff and that's what I'm working on. It's fun when I get bored. I can write new patches Oh, and I also have the read file syscall, which I keep posting every once in a while patches are public To move open read close into one syscall Interestingly, it doesn't always it isn't always faster a lot of workloads because It's lost in the noise. It's kind of interesting syscalls are not slow They're fast, but they're slow down a little bit, but they're still not that slow as you would think they were Linux in a safety critical environment Anybody who's ever flown a plane has been controlled by Linux in past decades. So it's in a safety critical environment today um, it runs the telecom systems. It runs this the stock markets that run satellites the um, it's been in the Bouncing thing that keeps super mega yachts from tipping over. So that's a safety critical environment It's been there for like 15 years So the Linux has been in the safety critical environments and has been automotive also as well in the head units for a long time Not in the head units. I don't know about where else it's going, but it's getting there for other parts of the car Nobody wants to write an operating system. There's one right applications to solve their problems So, yeah, we're there. Um, there are certifications of other paperwork that people are working on but that's independent of Linux itself I have three minutes left. Okay Um, which kernel feature turned my beard that gray all of them It's just me getting old and look, I don't know here you but I've not had it here a long time Um My source lts the security updates staying on lts And what is a good way to convince customers up to and ahead of lts or latest tables a way way to pick Um cherry picking always fails. I will guarantee you that I have audited lots of devices I did a famous talk where I showed four brand new phones that were bought within They were the latest top of long ones and I was trivial to root them all I did it on stage as a fun talk And it was rooted with a patch that had said this is a fix. Here's the problem Here's the reproducer in the commit itself and people who are cherry picking missed it um Google's published a bunch of results about how taking the lts fixes 90 of all the cvs before they're even done the other 10 of the cvs are security bugs were in out of shriek code that was in their kernels So every single thing that they were notified of being a security issue was fixed before they even knew about it In the lts releases. Um google's now requiring this for new devices that they pick up the lts releases Again cherry picking does not work. There's no reason to do that Just take the whole thing and do that because we fix known security issues every single week And we fix tons of unknown security issues every single week Again, we're fixing 30 a week for 30 at 35 a day Keeping on top of that and determining what is and is not a security thing is impossible. So just take them all So again, I've given presentations on it. You can look at what I gave last year and the before and just do that What do I think about open source security foundation? I do not know. I'm so sorry But I don't know what they're doing. I think it's a replacement for what ci I was doing I am not involved directly and so I can't tell you what to do there I can point you to the right people to talk to case cook. He's involved. Um, so alexander talk to him um I have no books coming out Um about crown development. Thanks to my story before it's not going to happen. So don't worry about that Um, just look at all the source code we have today and the ideas are still in the old books are still good Um What else I have They did not recover. They did not reconsider O'Reilly did not do that and they laid off all their editors. So I doubt that will ever happen Why is the kernel community not a friend of dbtk? I don't think you know that the same developers are part of the kernel community and part of dbtk So why are they? They are friends. They're the same people That's a not a question um That's about the collaboration process with intel and I will not I'll just do these last ones questions. Um Disclosure process intel is not doing well With disclosures I'm not happy It's not getting better. Um As proof the bluetooth problem Was intel it wasn't disclosed properly Anyway, I won't get into that Stable maintainership more maintainers bigger team bigger team be kind of productive is our sasha and I going too slow 35 batches a day if other people want to help that our active kernel maintainers I'm glad to take the help. We have a lot of things that would do that Um, I have a list of tasks to do if you want to help That's about it Um, the context regarding rust was in the context of drivers will be using core codes Let's see it actually work in drivers first come on What do you want us to say? How would it be used in core code if it can't even work in a driver? That's a leaf node on the tree as far as source code and not requiring no way is building up off of that baby steps, please And the current roadmap for iou ring. Um talk to yens. I think he's probably giving a talk Um, it's great stuff. It works really really fast and it's fun to use So talk to yens and he's given he read his emails and read the patches that are coming out from him He does a lot of good stuff And I think that is it. So thank you very much. If anyways, follow on questions. Just email me I look like I said I go about thousand emails a day. Might as well have interesting ones. Thanks a lot